Inconsistencies In Consistency Models: Better ODE Solving Does Not Imply Better Samples

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Vouitsis, Noël, Hosseinzadeh, Rasa, Ross, Brendan Leigh, Villecroze, Valentin, Gorti, Satya Krishna, Cresswell, Jesse C., Loaiza-Ganem, Gabriel
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915021646200832
author Vouitsis, Noël
Hosseinzadeh, Rasa
Ross, Brendan Leigh
Villecroze, Valentin
Gorti, Satya Krishna
Cresswell, Jesse C.
Loaiza-Ganem, Gabriel
author_facet Vouitsis, Noël
Hosseinzadeh, Rasa
Ross, Brendan Leigh
Villecroze, Valentin
Gorti, Satya Krishna
Cresswell, Jesse C.
Loaiza-Ganem, Gabriel
contents Although diffusion models can generate remarkably high-quality samples, they are intrinsically bottlenecked by their expensive iterative sampling procedure. Consistency models (CMs) have recently emerged as a promising diffusion model distillation method, reducing the cost of sampling by generating high-fidelity samples in just a few iterations. Consistency model distillation aims to solve the probability flow ordinary differential equation (ODE) defined by an existing diffusion model. CMs are not directly trained to minimize error against an ODE solver, rather they use a more computationally tractable objective. As a way to study how effectively CMs solve the probability flow ODE, and the effect that any induced error has on the quality of generated samples, we introduce Direct CMs, which \textit{directly} minimize this error. Intriguingly, we find that Direct CMs reduce the ODE solving error compared to CMs but also result in significantly worse sample quality, calling into question why exactly CMs work well in the first place. Full code is available at: https://github.com/layer6ai-labs/direct-cms.
format Preprint
id arxiv_https___arxiv_org_abs_2411_08954
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Inconsistencies In Consistency Models: Better ODE Solving Does Not Imply Better Samples
Vouitsis, Noël
Hosseinzadeh, Rasa
Ross, Brendan Leigh
Villecroze, Valentin
Gorti, Satya Krishna
Cresswell, Jesse C.
Loaiza-Ganem, Gabriel
Machine Learning
Artificial Intelligence
Although diffusion models can generate remarkably high-quality samples, they are intrinsically bottlenecked by their expensive iterative sampling procedure. Consistency models (CMs) have recently emerged as a promising diffusion model distillation method, reducing the cost of sampling by generating high-fidelity samples in just a few iterations. Consistency model distillation aims to solve the probability flow ordinary differential equation (ODE) defined by an existing diffusion model. CMs are not directly trained to minimize error against an ODE solver, rather they use a more computationally tractable objective. As a way to study how effectively CMs solve the probability flow ODE, and the effect that any induced error has on the quality of generated samples, we introduce Direct CMs, which \textit{directly} minimize this error. Intriguingly, we find that Direct CMs reduce the ODE solving error compared to CMs but also result in significantly worse sample quality, calling into question why exactly CMs work well in the first place. Full code is available at: https://github.com/layer6ai-labs/direct-cms.
title Inconsistencies In Consistency Models: Better ODE Solving Does Not Imply Better Samples
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2411.08954